With the considerable increase in ownership of motor vehicles, traffic crashes have become a challenge. This paper presents a study of naturalistic driving conducted to collect driving data. The experiments were performed on different road types in the city of Wuhan in China. The collected driving data were used to develop a near-crash database, which covers driving behavior, near-crash factors, driving environment, time, demographics, and experience. A new definition of near-crash events is also proposed. The new definition considers potential risks in driving behavior, such as braking pressure, time headway, and deceleration. A clustering analysis was carried out through a K-means algorithm to classify near-crash events based on their risk level. In addition, a mixed-ordered logit model was used to examine the contributing factors associated with the driving risk of near-crash events. The results indicate that ten factors significantly affect the driving risk of near-crash events: deceleration average, vehicle kinetic energy, near-crash causes, congestion on roads, time of day, driving miles, road types, weekend, age, and experience. The findings may be used by transportation planners to understand the factors that influence driving risk and may provide valuable insights and helpful suggestions for improving transportation rules and reducing traffic collisions thus making roads safer.

Driving risk varies substantially according to many factors related to the driven vehicle, environmental conditions, and drivers. This study explores the contributing historical factors of driving risk with hierarchical clustering analysis and the quasi-Poisson regression model. The dataset of the study was collected from two sources: naturalistic driving experiments and self-reports. The drivers who participated in the naturalistic driving experiment were categorized into four risk groups according to their near-crash frequency with the hierarchical clustering method. Moreover, a quasi-Poisson model was used to identify the essential factors of individual driving risk. The findings of this study indicated that historical driving factors have substantial impacts on individual risk of drivers. These factors include the total number of miles driven, the driver’s age, the number of illegal parking (past three years), the number of over-speeding (past three years) and passing red lights (past three years). The outcome of the study can help transportation officials, educators, and researchers to consider the influencing factors on individual driving risk and can give insights and provide suggestions to improve driving safety.

As a popular transportation mode in urban regions, taxis play an essential role in providing comfortable and convenient services for travelers. For the sake of tackling the imbalance between supply and demand, taxi demand forecasting can help drivers plan their routes and reduce waiting time and oil pollution. This paper proposes a deep learning-based model for taxi demand forecasting with multi-source data using Generative Adversarial Networks. Firstly, main features were extracted from multi-source data, including GPS taxi data, road network data, weather data, and points of interest. Secondly, Generative Adversarial Network, comprised of the recurrent network model and the conventional network model, is adopted for fine-grained taxi demand forecasting. A comprehensive experiment is conducted based on a real-world dataset of the city of Wuhan, China. The experimental results showed that our model outperforms state-of-the-art prediction methods and validates the usefulness of our model. This paper provides insights into the temporal, spatial, and external factors in taxi demand-supply equilibrium based on the results. The findings can help policymakers alter the taxi supply and the taxi lease rents for periods and increase taxi profit.

Identifying dangerous events from driving behavior data has become a vital challenge in intelligent transportation systems. In this study, we compared machine and deep learning-based methods for classifying the risk levels of near-crashes. A dataset was built for the study by considering variables related to naturalistic driving, temporal data, participants, and road geometry, among others. Hierarchical clustering was applied to categorize the near-crashes into several risk levels based on high-risk driving variables. The adaptive lasso variable model was adopted to reduce factors and select significant driving risk factors. In addition, several machine and deep learning models were used to compare near-crash classification performance by training the models and examining the model with testing data. The results showed that the deep learning models outperformed the machine learning and statistical models in terms of classification performance. The LSTM model achieved the highest performance in terms of all evaluation metrics compared with the state-of-the-art models (accuracy = 96%, recall = 0.93, precision = 0.88, and F1-measure = 0.91). The LSTM model can improve the classification accuracy and prediction of most near-crash events and reduce false near-crash classification. The finding of this study can benefit transportation safety in predicting and classifying driving risk. It can provide useful suggestions for reducing the incidence of critical events and forward road crashes.

Taxi trajectories reflect human mobility over the urban roads’ network. Although taxi drivers cruise the same city streets, there is an observed variation in their daily profit. To reveal the reasons behind this issue, this study introduces a novel approach for investigating and understanding the impact of human mobility patterns (taxi drivers’ behavior) on daily drivers’ profit. Firstly, a K-means clustering method is adopted to group taxi drivers into three profitability groups according to their driving duration, driving distance and income. Secondly, the cruising trips and stopping spots for each profitability group are extracted. Thirdly, a comparison among the profitability groups in terms of spatial and temporal patterns on cruising trips and stopping spots is carried out. The comparison applied various methods including the mash map matching method and DBSCAN clustering method. Finally, an overall analysis of the results is discussed in detail. The results show that there is a significant relationship between human mobility patterns and taxi drivers’ profitability. High profitability drivers based on their experience earn more compared to other driver groups, as they know which places are more active to cruise and to stop and at what times. This study provides suggestions and insights for taxi companies and taxi drivers in order to increase their daily income and to enhance the efficiency of the taxi industry.

Recently, hyperspectral image (HSI) classification has become a hot topic in the geographical images research area. Sufficient samples are required for image classes to properly train classification models. However, a class imbalance problem has emerged in hyperspectral image (HSI) datasets as some classes do not have enough samples for training, and some classes have many samples. Therefore, the performance of classifiers is likely to be biased toward the classes with the largest samples, and this can lead to a decrease in the classification accuracy. Therefore, a new deep-learning-based model is proposed for hyperspectral images generation and classification of imbalanced data. Firstly, the spectral features are extracted by a 1D convolutional neural network, whereas a 2D convolutional neural network extracts the spatial features and the extracted spatial features and spectral features are catenated into a stacked spatial–spectral feature vector. Secondly, an autoencoder model was developed to generate synthetic images for minority classes, and the image samples were balanced. The GAN model is applied to determine the synthetic images from the real ones and then enhancing the classification performance. Finally, the balanced datasets are fed to a 2D CNN model for performing classification and validating the efficiency of the proposed model. Our model and the state-of-the-art classifiers are evaluated by four open-access HSI datasets. The results showed that the proposed approach can generate better quality samples for rebalancing datasets, which in turn noticeably enhances the classification performance compared to the existing classification models.